Re: No console anymore. How to figure this out.

On 3/16/26 5:16 AM, Michael wrote:
> Your console device needs the nvidia-drm.ko module to be able to use
> nvidia's own kernel modesetting framebuffer functionality
> (modesetting=1).
>
> What I think is happening:  your EFI framebuffer is providing the
> initial console display during boot, then it switches over to nvidia-
> drm.
>
> I suspect your problem manifests at the handover from kernel to nvidia's
> module, or vice-versa, when you try to go back to a console.  However, I
> haven't used nvidia for decades now and have not kept up with their
> current driver's functionality to know with any certainty how this is
> meant to work.
>
> What does dmesg report at boot and thereafter when you switch to a
> console?
>
> Can you grep your running kernel's for 'DRM', to see if anything
> incompatible has been enabled/disabled?
>
> What does 'lsmod | grep nvidia' come up with?
>
> Also, 'modinfo nvidia-drm'?
>
> PS. I assume your have re-emerged your nvidia driver to match the kernel
> you are now running.


That's my thinking to, the last part.  I didn't think about the first 
part.  LOL  This is lsmod.  The drm part is loaded as is modeset.


root@Gentoo-1 / # lsmod
Module                  Size  Used by
nvidia_uvm           1814528  2
nvidia_drm            118784  40
nvidia_modeset       1875968  161 nvidia_drm
nvidia              110895104  2122 nvidia_uvm,nvidia_modeset
root@Gentoo-1 / #



So it appears the modules are loaded but not all are working.  I 
guess???  Here's the other info.



root@Gentoo-1 / # modinfo nvidia-drm
filename:       /lib/modules/6.12.31-gentoo/video/nvidia-drm.ko
version:        580.142
supported:      external
license:        MIT
srcversion:     7C9FF03229E79984059FE07
alias:          pci:v000010DEd*sv*sd*bc06sc80i00*
alias:          pci:v000010DEd*sv*sd*bc03sc02i00*
alias:          pci:v000010DEd*sv*sd*bc03sc00i00*
depends:        nvidia-modeset
name:           nvidia_drm
retpoline:      Y
vermagic:       6.12.31-gentoo SMP preempt mod_unload
parm:           modeset:Enable atomic kernel modesetting (1 = enable, 0 
= disable (default)) (bool)
parm:           fbdev:Create a framebuffer device (1 = enable (default), 
0 = disable) (bool)
root@Gentoo-1 / #



Yes, I did rebuild the nvidia drivers however, I'm still on the same 
kernel.  Been using the current kernel for ages now.  I think since 
shortly after I built this rig.  The kernel we were trying to adjust 
never would finish the make process, compile I guess.

Does anything there give a clue?  Could it be the nvidia drivers have a 
bug, in both versions in the tree?  My card is older but I can't be the 
only one that uses a card like that.  Of course, some people that use a 
GUI rarely switch to a console.  So, it could be a bug that no one else 
has noticed I guess.  Unless you or someone else has some more ideas to 
check.

In the meantime, I may switch to every other weekend to update.  It 
takes a while to get all my stuff running and open again when I reboot.  
I try to set aside 30 minutes or so.  Sometimes, it takes close to 
that.  It's not the booting itself, it's me.  ;-)

Thanks.

Dale

:-)  :-)